Talk:Install Zenoss

From Zenoss Wiki
Revision as of 14:43, 16 August 2012 by Makemorebeer (Talk | contribs) (Centos 6.3 failed to install at Zenoss: new section)

Jump to: navigation, search


If you did a manual install of Centos 6.3, the firewall will be enabled by default, and you will not be able to connect to the Zenoss web page on port 8080 until you disable it:

  1. service iptables save
  2. service iptables stop
  3. chkconfig iptables off

Updated end-of-install docs

I have tweaked the auto-deploy script to provide firewall info after it completes successfully. See

Command failure: /sbin/service zenoss start


I have try to install Zenoss with the auto-script, but I have this issue:

Starting MySQL SUCCESS! Zenoss not initialized. Performing first-boot initialization... Fresh install pre steps Please confirm that the JRE in the PATH is Oracle's version. Command failure: /sbin/service zenoss start

Thank you for your help.

Edit: I am on a Centos 6.3

This should now be fixed - you probably had java-openjdk installed. The updated version of the script should now remove it first so that it will find the Oracle JRE. -DR

Command failure: /sbin/service zenoss start - 2nd observation

Failes for me to!

Wget failure zenoss 4.2.0 packages from

[root@zenoss-master ~]# ./ Ensuring This server is in a clean state before we start Ensuring Zenoss RPMs are not already present Disabling SELinux... Downloading zenoss-4.2.0.el6.x86_64.rpm... --2012-08-08 14:42:24-- Resolving Connecting to||:80... connected. HTTP request sent, awaiting response... 404 Not Found 2012-08-08 14:42:24 ERROR 404: Not Found.

Command failure: wget -N

Try again - I have confirmed the URL is correct and this path exists. This may have been due to a sourceforge quirk at the time. - DR

mysql service

on 64 bit centos 6.3 I had to modify installation script so it started mysqld service, because mysql service didn't exist. I also noticed a bunch of notifications for setting env variables trying to access /usr/local/zenoss, while installation was put in /opt/zenoss. But it works anyway.

Centos 6.3 base install mysql-libs

I did a base install of Centos 6.3 using the netinstall image and choosing minimal. I did not install any other packges except wget. When I ran the Auto Deploy script it said it couldn't continue because I had an older version of the mysql-libs library installed. I used yum to remove them which also removed postfix. After doing this the Auto Deploy script ran fine, however now I don't have postfix installed, if I try to do a yum to install it again it wants to install mysql-libs, and I"m not sure if this will break the Zenoss install.

-- I found that if I do a yum remove mysql-libs it removes many packages, however, if I followed it up with a "yum install sysstat" (one of the packages it removed) it added them all back, except mysql-libs

-- I replace postfix with sendmail, thus #yum install sendmail; yum remove mysql-libs Then only mysql-libs and postfix removed, not cron packages as well.

I have created the following bug to track this issue:

Centos 6.2 ZODB installation issue

Running the script gives the following error on a fresh install of CentOS release 6.2:

  zenoss.x86_64 0:4.2.0-1586.el6                                                                                    

Configuring and Starting some Base Services
Starting memcached: 
Starting snmpd: 
Zenoss not initialized.  Performing first-boot initialization...
Fresh install pre steps
Checking RRDtool version >= 1.4.7: [ OK ] 1.4.7
ERROR 1548 (HY000) at line 6 in file: '/opt/zenoss/Products/ZenUtils/relstorage/mysql/001.sql': Cannot load from mysql.proc. The table is probably corrupted
Dropping database: zodb_session
Creating database: zodb_session
Applying schema version: 1
Failed to create ZODB session database
Command failure: /sbin/service zenoss start
[root@localhost zenoss]#

Centos 6.3 rabbit-mq issue

[root@Muon-Z ~]# chkconfig iptables off [root@Muon-Z ~]# service rabbitmq-server start Starting rabbitmq-server: FAILED - check /var/log/rabbitmq/startup_{log, _err} rabbitmq-server.

in the /var/log/rabbitmq/ startup_log I get

Activating RabbitMQ plugins ... 0 plugins activated:

ERROR: epmd error for host "Muon-Z": address (unable to establish tcp connection)

How is this corrected?

The solution is by adding the server name in the hosts file


CentOS 6.3 Fresh Install

Attempting a fresh install using the automated script. Minimal server CentOS installation with only wget added to allow download of the script. Script installs everything until the zenoss start portion:

Configuring and Starting some Base Services
Starting memcached:                                        [  OK  ]
Starting snmpd:                                            [  OK  ]
Zenoss not initialized.  Performing first-boot initialization...
Fresh install pre steps
Checking RRDtool version >= 1.4.7: [ OK ] 1.4.7
grep: Unmatched [ or [^
Adding RabbitMQ user:
grep: Unmatched [ or [^
Adding RabbitMQ vhost:
Setting RabbitMQ permissions for user:
Usage: zeneventserver-create-db [options]

zeneventserver-create-db: error: option --dbport: invalid integer value: '\x1b[H\x1b[2J3306'
Failed to create ZODB session database
Command failure: /sbin/service zenoss start

It appears to be a problem with the zenoss pre install script, but I am unsure.


Centos 6.3 failed to install at Zenoss

I ran into some of the problems listed, such as the mysql-libs, and a couple other dependancy issues, but got through them all, but this one doesn't give me much to work with and since I'm not real experianced with linux i'm not sure how to work around it.

Installing Zenoss Loaded plugins: fastestmirror, refresh-packagekit, security Loading mirror speeds from cached hostfile epel/metalink | 11 kB 00:00

* base:
* epel:
* extras:
* updates:

Setting up Local Package Process Cannot open: zenoss-4.2.0.el6.x86_64.rpm. Skipping. Nothing to do Configuring and Starting some Base Services error reading information on service memcached: No such file or directory Command failure: /sbin/chkconfig memcached on